US 12,277,941 B2
Parameter encoding and decoding
Alexandre Bouthéon, Erlangen (DE); Guillaume Fuchs, Erlangen (DE); Markus Multrus, Erlangen (DE); Fabian Küch, Erlangen (DE); Oliver Thiergart, Erlangen (DE); Stefan Bayer, Erlangen (DE); Sascha Disch, Erlangen (DE); and Jürgen Herre, Erlangen (DE)
Assigned to Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V., Munich (DE)
Filed by Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V., Munich (DE)
Filed on Dec. 14, 2021, as Appl. No. 17/550,931.
Application 17/550,931 is a continuation of application No. PCT/EP2020/066456, filed on Jun. 15, 2020.
Claims priority of application No. 19180385 (EP), filed on Jun. 14, 2019.
Prior Publication US 2022/0108707 A1, Apr. 7, 2022
Int. Cl. G10L 19/008 (2013.01); G10L 19/08 (2013.01); H04S 3/02 (2006.01)
CPC G10L 19/008 (2013.01) [G10L 19/08 (2013.01); H04S 3/02 (2013.01); H04S 2400/01 (2013.01); H04S 2400/03 (2013.01)] 34 Claims
OG exemplary drawing
 
1. An audio synthesizer for generating a synthesis signal from a downmix signal, the synthesis signal comprising a at least three synthesis channels, the audio synthesizer comprising:
an input interface configured for receiving the downmix signal, the downmix signal comprising a plural number of downmix channels and side information, the side information comprising channel level and correlation information of an original signal, the original signal comprising a plural number of original channels;
a prototype signal calculator configured for calculating a prototype signal from the downmix signal, the prototype signal comprising the number of synthesis channels, the prototype signal calculator being configured to apply a prototype matrix to the downmix signal to obtain the prototype signal; and
a synthesis processor configured for generating the synthesis signal by applying, to the prototype signal, at least one mixing rule in form of a matrix, the mixing rule being obtained from:
channel level and correlation information of the original signal, the channel level and correlation information being written in the bitstream; and
covariance information of the downmix signal,
wherein the audio synthesizer is configured to reconstruct a target version of the covariance information based on an estimated version of the original covariance information; reported to the number of synthesis channels, wherein the audio synthesizer is configured, in order to reconstruct the target version of the covariance information, to:
acquire the estimated version of the original covariance information by applying, to the covariance information of the downmix signal, the prototype matrix for calculating a prototype signal, so as to report the estimated version of the original covariance information to the number of synthesis channels,
normalize first values of the estimated version of the original covariance information reported to the number of original channels;
retrieve further normalized values of the original covariance information from the channel level and correlation information of the original signal written in the side information, and assign the further normalized values of the original covariance information to channels of the synthesis channels, thereby reporting the further normalized values of the original covariance information to the number of original channels;
denormalize the first normalized values and the further normalized values, to acquire a denormalized version of the original covariance information reported to the number of original channels
thereby retrieving the target version of the covariance information, to thereby derive the mixing rule using the target version of the covariance information,
so that the synthesis processor generates the synthesis signal using the prototype signal and the at least one mixing rule.